The use of vehicle navigation systems is now ubiquitous. Navigation systems typically track the location of a vehicle and display the location along with map data in a human-readable format. Additionally, navigation systems typically output visual and audio directions associated with a mapped route. However, the directions output by conventional navigation systems are typically based on distances and street names. For example, a driver may be instructed to “turn right onto Main Street in 100 feet.” In certain instances, a user may have difficulty in maintaining a desired course based upon these conventional instructions. For example, a street name may not be clearly marked, thereby increasing a likelihood that a user turns onto the wrong street. Accordingly, there is an opportunity for improved systems and methods for generating descriptive outputs associated with navigation directions.